Multi-omic analysis identifies BACH2 transcription factor as key epigenetic and transcriptional repressor driving drug-tolerance to targeted therapy in EGFR-mutant lung adenocarcinoma

RATIONALE Drug resistance poses a significant challenge for EGFR-mutant lung cancer patients treated with osimertinib, a 3rd-generation EGFR tyrosine kinase inhibitor. Non-mutational mechanisms, particularly epigenetic alterations in drug-tolerant persister (DTP) cells, enable survival under lethal drug pressure and seed resistance. A deeper understanding of these mechanisms is required to evaluate whether targeting epigenetic changes in DTP cells may represent a promising therapeutic strategy. METHODS We conducted a preclinical multi-omic study to map the epigenetic and transcriptional changes driving DTP in EGFR-mutant lung cancer. Using PC9-DTP cell line exposed to osimertinib for 14 days, we employed a combination of single-cell and bulk omics assays to profile genome-wide DNA methylation, chromatin accessibility, and transcriptomic landscapes defining the DTP state. Top epigenetic regulators were nominated from integrative analyses, and public datasets from other EGFR-mutant lung cancer models were analyzed for validation. RESULTS DTP cells displayed global DNA hypermethylation across the vast majority (88.7%) of differentially methylated regions. Hypermethylation in cis-regulatory regions were strongly associated with closed chromatin and reduced transcription (R=-0.64, p<2.2e-16), reflecting coordinated epigenetic silencing. Gene regulatory network analysis and in silico perturbation, leveraging paired transcriptomic and chromatin features from single-cell multi-omic assay, identifies BACH2 - transcriptional and epigenetic repressor - as the top repressive transcription factor in DTP-specific networks. Additionally, chromatin remodeler HMGA1 was among the target genes most strongly repressed by BACH2. Based on motif enrichment analysis, BACH2-specific DNA-binding motifs were enriched in hypermethylated (p<1e-4) and closed chromatin regions (p<1e-6), accompanying the downregulation of corresponding target genes, including HMGA1. Furthermore, pathway analysis of BACH2-repressed target genes revealed enrichment in oxidative phosphorylation and ATP synthesis, indicating a metabolic shift. Finally, analyzing public transcriptomic datasets from other cell line models, patient-derived organoids (PDOs), xenografts (PDXs), and patient tumor samples confirmed consistent BACH2 upregulation in DTP and residual disease state, as well as downregulation of HMGA1 and other target genes. CONCLUSION These findings position BACH2 as a key contributor of epigenetic silencing and metabolic reprogramming in osimertinib-induced DTP state. Overall, our multi-omic approach provides high-resolution insight into drug-tolerance mechanisms and nominates BACH2 as a priority target, potentially informing novel therapeutic strategy for improved patient outcome.